Our studies of American Indian youth, along with those of others, have demonstrated the substantially increased risk among native youth for substance use and other related problem behaviors. Our 25-year tracking of adolescent substance use among American Indians shows that native youth have reported consistently higher rates of substance use than other U.S. ethnic minority groups and White non-Hispanic youth. Because of their increased risk, it is essential to further understand the nature and course of substance use among this population of youth. The overall goal of this center project will be to examine the growth and the trajectories of Indian adolescent substance use over an extended period of time, determining antecedents and outcomes associated with substance use change over time. Considerable research has investigated latent growth and latent class trajectories of alcohol use among youth in general. But such investigation has not been conducted among native youth. Essentially no past research has investigated latent class trajectories for substance use other than alcohol among adolescent youth. We plan to investigate these questions investigating American Indian adolescents using a unique longitudinal sample of American Indian youth and their families. The American Indian Research Center at the University of Washington has conducted a prospective study of American Indian adolescent substance use. The study enrolled both adolescents and their families and conducted annual assessments each year over a nine-year period. A 96% retention rate was achieved over the course of the study resulting in usable data on approximately 500 families. An extensive range of variables were measured including demographic, psychological, behavioral, social, and family domains. Adolescent subjects were followed from a mean age of 11 through a mean age of 19, the age of highest risk for development of alcohol and drug-related adverse consequences. While some use has been made of this data, it has not been analyzed using newer statistical procedures that will allow us to assess growth models of substance use and latent classes of trajectories using growth-mixture models. A number of hypotheses will be testable including identification of substance use trajectories, detection of predictors and consequences of different trajectories, and timing and interactions of specific etiologic factors. [unreadable] [unreadable]